Doctoral Dissertation Mean Field Equilibrium Asset Pricing Models With Exponential Utility

2026

Read Doctoral Dissertation Mean Field Equilibrium Asset Pricing Models With Exponential Utility (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-22058.txt.

Introduction

Sekine's thesis builds asset pricing models for markets with many unlike agents. The tool is mean field game theory, and markets are incomplete. Agents differ in wealth, risk aversion, and unspanned liability, and all use exponential utility. Sekine derives the equilibrium risk premium from inside the model. He characterizes equilibrium with a new mean field backward stochastic differential equation. He proves solutions exist and clear the market for large populations. Later parts add consumption with habit formation, plus markets where agents infer the premium from prices.

Catalog record

YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE19368
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.22058

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Doctoral Dissertation Mean Field Equilibrium Asset Pricing Models With Exponential Utility Abstract Acknowledgement First and foremost, I would like to express my deepest gratitude to Associate Professor Masaaki Fujii for his long-term guidance and support. His passion has taught me the depth and joy of research and his earnest attitude has been and always will be a guiding principle in my life. Over the past seven years, he has devoted countless hours to mentoring me and providing valuable feedback on my papers, which has encouraged me to think critically and pushed me to achieve my best. I would also like to express my sincere gratitude to my parents, Tsuyoshi and Shizuka, for their thoughtful advice and understanding, which enabled me to enjoy my research life and keep challenging myself. I extend my heartfelt thanks to my dear wife, Mina, for her constant love and support since my undergraduate days. I would like to thank my dear friends and relatives for their warm support and encouragement. I am truly fortunate to have such wonderful people by my side. This research is supported by the Grant-in-Aid for JSPS Fellows (Grant Number JP23KJ0648). Tokyo, Japan November 2025 Masashi Sekine Abstract This thesis develops equilibrium asset pricing models in incomplete markets with a large number of heterogeneous agents, using mean field game theory. The contributions of the thesis are summarized below: 1. It proposes novel models in asset pricing theory and characterizes the equilibrium using a novel form of equation. 2. It examines the properties of the relevant equation and derives the risk premium that achieves market clearing under certain conditions. 3. It provides several extensions for general settings. The thesis begins with a brief introduction and review of the field, followed by three models of mean field asset pricing theory. The first model is a theoretical model that endogenously derives the equilibrium risk premium of securities in a market with heterogeneous agents. The agents are assumed to have exponential-type preferences and to be heterogeneous in the initial wealth, the coefficient of absolute risk aversion, and the stochastic terminal liability, which is unspanned by the security prices. The study begins by solving the optimal investment problem using the optimal martingale principle. The market equilibrium is characterized by the solution of a novel mean field backward stochastic differential equation (BSDE), whose driver has quadratic growth both in the stochastic integrands and in their conditional expectations. The existence of a solution to such an equation is proved under several conditions. The equilibrium risk-premium process is characterized in terms of the solution to this mean field BSDE and is shown to clear the market in the large population limit. The second model is an extension of the first model; it considers agents’ consumption and incorporates habit formation. The incorporation of habit formation relaxes the assumption of time-separable utility functions by making utility depend not only on the current level of consumption but also on the agents’ accumulated stock of past consumption. In order to characterize the market-clearing equilibrium, a mean field BSDE, which has a similar form to that of the first model, is derived. Its well-posedness and asymptotic behavior are then examined. Additionally, an exponential quadratic Gaussian (EQG) reformulation of the asset pricing model is introduced, in which an equilibrium solution is obtained in a semi-analytic form. The third model extends the first model to the problem of a partially observable market with a large number of heterogeneous agents. This model assumes that agents can only observe stock prices and must infer the risk premium from these observations when determining trading strategies. The equilibrium risk premium in such a market is characterized in terms of a solution to the mean field BSDE. Specifically, the solution to the mean field BSDE can be expressed semi-analytically by employing the EQG framework. The equilibrium risk-premium process is then constructed endogenously using Kalman-Bucy filtering theory. In addition, a simple numerical simulation is included to visualize the dynamics of the market model. Keywords Asset pricing, Backward stochastic differential equation, Exponential utility, Market clearing, Mean field game, Stochastic optimal control Mathematics Subject Classification 49N80, 60H10, 91B51, 91G10, 93E11 JEL Classification C61, C73, D52, G11, G12 Chapter 1 Introduction Financial markets have enabled smooth funding and effective risk allocation among participants in contemporary society. Mathematical finance and financial economics are fields of study that mathematically analyze various phenomena in financial markets, and they have developed rapidly over the past half-century commencing with pioneering work such as Markowitz [ portfolio_selection ] and Black & Scholes [ black-scholes ] . Specific areas of this study include the valuation of financial derivatives, risk management of investment assets, and the formulation of optimal investment strategies, all of which are widely applied in financial practice. Asset pricing theory also plays a crucial role in this field. The fundamental objective of the theory is to establish the equilibrium price at which the supply of assets matches the demand. The capital asset pricing model (CAPM), proposed by Sharpe [ w_sharpe ] , Lintner [ j_lintner ] , and Mossin [ mossin ] , is the most commonly known model, and these theories provide deep insights into how asset prices are determined in the financial market. Various economic phenomena can be explained by the interaction of a large number of individuals who make decisions in an attempt to maximize their own profits. This situation can be analyzed within the framework of a “multiplayer game.” Financial markets are, of course, no exception, since they are places where a large number of investors trade on a daily basis, making tactical decisions to execute their long-term strategic goals. In the classical literature of mathematical finance, particularly in models of derivative pricing, optimal investment problems, and portfolio risk management, market environments are typically assumed to be exogenous. These models can describe how investors behave in a given market environment. However, market environments, especially security prices, are essentially determined by the balance of investors’ buy and sell orders, and it is therefore desirable to consider a theoretical model that can express the market environments endogenously. Moreover, standard asset pricing models often assume the existence of representative investors. While such an assumption has the advantage of analytical simplicity, it cannot describe situations in which each investor adopts a different investment strategy based on their own information, and equilibrium prices are determined through their interactions. To overcome these issues, we need a theoretical model that can capture the financial market endogenously within the model and take into account the competitive or cooperative behavior of heterogeneous investors. In other words, a model that views various phenomena of financial markets as a multiplayer game is necessary. If such a model is developed, it becomes possible to analyze how investor interactions influence the formation of security prices and how market parameters affect the equilibrium state. In particular, it can analyze how the existence of agents’ idiosyncrasies affects the price formation as a whole. However, a typical approach to multiplayer games usually becomes mathematically intractable since it leads to a strongly coupled optimization problem. The solution of such a problem can hardly be obtained analytically or even numerically, especially when the number of players is large, such as participants in financial markets.

Read the complete text